RE: Re: $1

[Odud]

I should have given an example of /x, here's how my version of your code that I used to try out my solution looks:

while (<>){
    if(/(
        [0-3]?[0-9](?:th|st|nd|rd)?\s+
        
        (?:Jan(uary)?|Feb(ruary)?|Mar(ch)?|Apr(il)?|May|Jun(e)?|
        
        Jul(y)?|Aug(ust)?|Sep(tember)?|Oct(ober)?|Nov(ember)?|Dec(embe
+r?))\s+
        
        [0-9]?[0-9]?[0-9][0-9])
        
        /ix) {
    print "$1\n";
    }
}

[Ted Nitz]

I like that regex more, but there are still a few minor changes I would make.

while (<>){
    if(/
        (
            [0-3]?[0-9](?:th|st|nd|rd)?\s+    # Get day
            (?:            # Get month
                Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|
                Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|
                Aug(?:ust)?|Sep(?:tember)?|Oct(?:ober)?|
                Nov(?:ember)?|Dec(?:ember)?
            )\s+
            [0-9]{2,4}            # Get Year
        )
    /ix) {
    print "$1\n";
    }
}

The (?:) things are just like normal ()'s except they don't capture to a variable, they just do grouping, helps the interpreter optimize the regex.

The {x,y} means require atleast x of the previous item, and at most y. For refrence you can also say {x} which means require x of them, or {x,} which means atleast x.
-Ted

[Odud]

Oh I agree entirely - it was a just a quick solution to the problem, not an attempt at an optimal one. There's a date parser module available through CPAN which I would normally use. I don't know if it copes with 2nd, 4th, etc. though.